1. Field of the Invention
The present invention relates to a computer program product, system, and method for recording backup information for backed-up data items in a data item list.
2. Description of the Related Art
With the rapid development of information technology, there is an increasingly huge amount of data to be stored and processed. The amount of data of many applications may be several hundred gigabytes (Gb) or even more. In general, an application data set usually includes many data items, and has an index file which will record information of respective data items in the data set. The information of data items may include: created time, size, and storage location of data items, and so on. In the case of a relatively huge application data set, the size of index file is also more than 1 Gb.
On the other hand, due to requirements of security, a variety of data often needs backup so as to prevent the risk of data loss. Typically, in order to complete data backup, the application data set is backed up into a server with mass storage space by means of a backup tool. Such backup may be regularly performed periodically, and alternatively may also be performed according to user settings. In the process of data backup, the backup tool backups the application data set together with its index file as a data packet into the server, and the server will store such data packet as one metadata. For each of such metadata, the server will create a description file, in which created time, size, and storage location of the metadata, and so on, are recorded. As a result, the server may record basic information of respective backup versions of the application data set according to backup time. Based on the basic information, a user may query and restore a particular version of the application data set according to the backup time.
However, in many cases, a user simply wants to query a particular data item in the application data set. Because the above backup and query mechanism cannot provide query for a particular data item, and the user usually cannot know for sure the created time or backup time of the data item to be queried, the user can only query with respect to respective versions of the data set. Specifically, the user needs to restore the index file of each version of the data set from the server, obtain information of data items according to the content of the index file, and then find a data item to be queried based on the information. However, in many cases, the backup of the data set is set to be regularly performed in a very short period, and especially in the case of incremental backup, a backup cycle is often at half an hour or even less. This results in a large number of different versions of the data set and the index files thereof. As mentioned earlier, in the case of a large data set, the size of an index file is often more than 1 G. Therefore, restoring each version of the index file from the server will cost a lot of time. Moreover, after the index file is restored, the user himself is still required to analyze respective index files one by one to find the data item to be queried. Undoubtedly, the above query process consumes a lot of time and effort of the user.
Therefore, there is a need for a solution that can improve backup and query of data items.